CH 9

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

CHAPTER 9

The waiting time in minutes before being served in a local post office is observed for 50 randomly
chosen customers: 2.1 0.5 3.6 1.4 2.0 0.8 0.4 4.2 3.5 2.5 4.8 2.8 1.9 1.2 3.2 1.6 2.5 2.4 1.9 2.0 3.5 5.2
3.1 1.6 1.5 1.9 2.4 2.7 2.1 1.8 4.6 3.8 1.5 4.5 3.9 5.5 2.5 3.8 5.0 4.6 2.1 2.8 1.6 3.8 4.2 3.5 5.2 4.8 3.9
2.6

(a) Test for normality using α= 0.05. What inferences can you draw?

(b) Estimate the mean and standard deviation of the waiting times.

(c) If the goal of the post office is for the waiting time not to exceed 4 minutes, find the capability
indices CPU and Cpk and comment on these values. Assuming normality, what proportion of the
customers, if any, will have to wait for more than 4 minutes?

CODE:
library(ggplot2)

wait_time <- c(33.2, 29.4, 36.5, 38.1, 30.0, 29.1, 32.2, 29.5, 36.0, 31.5, 34.5, 33.6, 27.4, 30.4,

28.4, 32.6, 30.4, 31.8, 29.8, 34.6, 30.7, 31.9, 32.3, 28.2, 27.5,34.9, 32.8, 27.7, 28.4, 28.8,

30.2, 26.8, 27.8, 30.5, 28.5, 31.8, 29.2, 28.6, 27.5, 28.5, 30.8, 31.8, 29.1, 26.9, 34.2, 33.5,

27.4, 28.5, 34.8, 30.5)

ggplot(mapping = aes(sample = wait_time)) + stat_qq_point(size = 2,color = "red") + stat_qq_line(color="green") +


xlab("Call Waiting Time") + ylab("Percent")

RESULT:
CHAPTER 9

COMMENTS:

From Anderson darling test the p-value is 0.129 which is greater then alpha = 0.05. So we accept the
null hypothesis of normality.

Sample mean = 2.906 minutes

Sample std. dev = 1.327 minutes.


CHAPTER 9

Consider the data on call waiting time of customers in a call centre (Exercise 5-9). The call centre has
set a goal of waiting time not to exceed 35 seconds.

(a) Test to see (using α =0.05) if conducting capability analysis using normal distribution is
appropriate.

(b) If not, consider a Box–Cox transformation and conduct capability analysis. Report appropriate
capability indices and the percentage nonconformance.

(c) Consider conducting capability analysis using a Weibull distribution. Comment on the results.

(d) What are the drawbacks of conducting a capability analysis using the normal distribution in this
example?

CODE:
wait_time <- c(33.2, 29.4, 36.5, 38.1, 30.0, 29.1, 32.2, 29.5, 36.0, 31.5, 34.5, 33.6, 27.4, 30.4, 28.4, 32.6, 30.4, 31.8, 29.8,
34.6, 30.7, 31.9, 32.3, 28.2, 27.5,34.9, 32.8, 27.7, 28.4, 28.8, 30.2, 26.8, 27.8, 30.5, 28.5, 31.8, 29.2, 28.6, 27.5, 28.5, 30.8,
31.8, 29.1, 26.9, 34.2, 33.5, 27.4, 28.5, 34.8, 30.5)

ggplot(mapping = aes(sample = wait_time)) + stat_qq_point(size = 2,color = "red") +

stat_qq_line(color="green") + xlab("Call Waiting Time") + ylab("Percent")

RESULT:
CHAPTER 9

CODE:
wait_time <- c(33.2, 29.4, 36.5, 38.1, 30.0, 29.1, 32.2, 29.5, 36.0, 31.5, 34.5, 33.6, 27.4, 30.4, 28.4, 32.6, 30.4,
31.8, 29.8, 34.6, 30.7, 31.9, 32.3, 28.2, 27.5,34.9, 32.8, 27.7, 28.4, 28.8, 30.2, 26.8, 27.8, 30.5, 28.5, 31.8, 29.2,
28.6, 27.5, 28.5, 30.8, 31.8, 29.1, 26.9, 34.2, 33.5, 27.4, 28.5, 34.8, 30.5)

hist(wait_time, col = "yellow") new_wait_time = 1/(wait_time)^2 hist(new_wait_time)


shapiro.test(new_wait_time) boxcox(wait_time, objective.name = "Log-Likelihood", optimize = TRUE)

boxcox.list <- boxcox(wait_time) plot(boxcox.list) plot(boxcox.list, plot.type = "Q-Q Plots", same.window =


FALSE)

RESULT:
CHAPTER 9

CODE:
Shapiro-Wilk normality test

data: new_wait_time

W = 0.97135, p-value = 0.2624

$lambda

[1] -2

$data [1] 33.2 29.4 36.5 38.1 30.0 29.1 32.2 29.5 36.0 31.5 34.5 33.6 27.4 30.4 28.4 32.6 30.4 31.8
29.8 34.6 30.7 31.9 32.3

[24] 28.2 27.5 34.9 32.8 27.7 28.4 28.8 30.2 26.8 27.8 30.5 28.5 31.8 29.2 28.6 27.5 28.5 30.8 31.8
29.1 26.9 34.2 33.5

[47] 27.4 28.5 34.8 30.5

$sample.size

[1] 50

$data.name

[1] "wait_time"
CHAPTER 9

CODE:

eweibull(wait_time, method = "mle")

curve(dweibull(x, shape=11.03378, scale = 32.08403), from=20, to=40,

main = 'Weibull Distribution for Wait Time',

ylab = '',

lwd = 2,

col = 'green')

Result:

$distribution

[1] "Weibull"

$sample.size

[1] 50

$parameters

shape scale

11.03378 32.08403

$n.param.est

[1] 2

$method

[1] "mle"

$data.name

[1] "wait_time"

$bad.obs

[1] 0

attr(,"class")

[1] "estimate"
CHAPTER 9

RESULTS:
CHAPTER 9

mple?

You might also like